OpenAI Custom Chip Signals AI Hardware Race
Key Questions
What is OpenAI developing in custom AI hardware?
OpenAI is developing its Jalapeño custom silicon chip aimed at superior performance per watt, with plans for gigawatt-scale deployment by 2026.
How might OpenAI's chip affect NVIDIA and other hardware companies?
The chip directly challenges NVIDIA's market position and influences investment theses around Broadcom, Celestica, and other data center hardware players.
What is NVIDIA's response to competition in AI hardware?
NVIDIA is countering with optimized models such as GLM-5.2 (753B MoE) designed for its Blackwell GPUs.
What does Micron's market performance indicate about AI semiconductor demand?
Micron's $1.4T market cap underscores the strong ongoing demand for semiconductors driven by AI growth.
What new infrastructure is emerging for AI agents?
Identity and trust infrastructure for AI agents, such as Proof's x401 protocol, is emerging as a hardware-adjacent category.
OpenAI is developing its own custom silicon (Jalapeño chip) targeting superior performance per watt, with deployment at gigawatt scale by 2026. This directly impacts NVIDIA's market position and investment theses around Broadcom, Celestica, and other data center hardware players. NVIDIA counters with optimized models like GLM-5.2 (753B MoE) for Blackwell GPUs. Micron's $1.4T market cap underscores semiconductor demand. Identity and trust infrastructure for AI agents (Proof's x401 protocol) emerges as a new hardware-adjacent category.